library(magrittr)
library(tidyverse)
library(Seurat)
library(readxl)
library(cowplot)
library(colorblindr)
library(viridis)
library(progeny)
library(destiny)

theme_cowplot2 <- function(...) {
  theme_cowplot(font_size = 12, ...) %+replace%
    theme(strip.background = element_blank(),
          plot.background = element_blank())
}
theme_set(theme_cowplot2())
coi <- params$cell_type_super
cell_sort <- params$cell_sort
cell_type_major <- params$cell_type_major
louvain_resolution <- params$louvain_resolution
louvain_cluster <- params$louvain_cluster

1 Cluster markers

1.1 Major T.super markers for cell assign

### load all data ---------------------------------
source("_src/global_vars.R")

# seu_obj <- read_rds(paste0("/work/shah/isabl_data_lake/analyses/16/52/1652/celltypes/", coi, "_processed.rds"))
seu_obj <- read_rds(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/outs_pre/", coi, "_seurat_", louvain_resolution, ".rds"))

myfeatures <- c("umapharmony_1", "umapharmony_2", "sample", louvain_cluster, "doublet", "nCount_RNA", "nFeature_RNA", "percent.mt", "doublet_score")

plot_data_wrapper <- function(cluster_res) {
  cluster_res <- enquo(cluster_res)
  as_tibble(FetchData(seu_obj, myfeatures)) %>% 
    left_join(meta_tbl, by = "sample") %>% 
    rename(cluster = !!cluster_res) %>% 
    mutate(cluster = as.character(cluster),
           tumor_supersite = ordered(tumor_supersite, levels = names(clrs$tumor_supersite)))
}

plot_data <- plot_data_wrapper(louvain_cluster)

1.2 Subtype currated markers

helper_f <- function(x) ifelse(is.na(x), "", x)

markers_v6_super[[coi]] %>% 
  group_by(subtype) %>% 
  mutate(rank = row_number(gene)) %>% 
  spread(subtype, gene) %>% 
  mutate_all(.funs = helper_f) %>% 
  formattable::formattable()

rank	CD4.T.dysfunctional	CD4.T.naive	CD4.T.reg	CD8.T	Cycling.T.NK	MT.high.T.NK	NK.CD56	NK.Cytotoxic
1	CD4	CCR7	CD4	CCL4	CDC20	IGKC	CD63	ADGRG1
2	CD40LG	CD4	FOXP3	CD8A	CDK1	MALAT1	CD7	CX3CR1
3	CTLA4	IL7R	IL2RA	CD8B	MKI67	MIAT	FCER1G	FCER1G
4	CXCL13	KLF2	TNFRSF4	CRTAM	PTTG1	MT-ND6	GNLY	FCGR3A
5	FKBP5	TCF7	TRAC	GZMA	TOP2A	MTRNR2L12	KLRC1	FGFBP2
6	IL6ST			GZMB		XIST	KLRD1	GNLY
7	ITM2A			GZMK			KLRF1	GZMH
8	MAF			GZMM			KRT81	IGFBP7
9	NMB			IFNG			KRT86	KLRD1
10	NR3C1			LAG3			NCAM1	KLRF1
11	PDCD1			MT1E			NKG7	NKG7
12	TNFRSF4			MT1X			TYROBP	PLAC8
13	TOX2			MT2A			XCL1	PLEK
14	TSHZ2			PTMS			XCL2	PTGDS
15				TRGC2				SPON2
16								TYROBP

1.3 Subtype cluster markers

# marker_tbl <- read_tsv(paste0("/work/shah/isabl_data_lake/analyses/16/52/1652/celltypes/", coi, "_markers.tsv")) %>% 
#   filter(resolution == louvain_resolution)
marker_tbl <- read_tsv(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/outs_pre/", coi, "_markers_", louvain_resolution, ".tsv"))

## Hypergeometric test --------------------------------------

test_set <- marker_tbl %>% 
  group_by(cluster) %>% 
  filter(!str_detect(gene, "^RPS|^RPL")) %>% 
  slice(1:50) %>% 
  mutate(k = length(cluster)) %>% 
  ungroup %>%
  select(cluster, gene, k) %>% 
  mutate(join_helper = 1) %>% 
  group_by(cluster, join_helper, k) %>%
  nest(test_set = gene)

markers_doub_tbl <- markers_v6 %>% 
  enframe("subtype", "gene") %>% 
  filter(!(subtype %in% unique(c(coi, cell_type_major)))) %>% 
  unnest(gene) %>% 
  group_by(gene) %>% 
  filter(length(gene) == 1) %>% 
  mutate(subtype = paste0("doublet.", subtype)) %>% 
  bind_rows(tibble(subtype = "Mito.high", gene = grep("^MT-", rownames(seu_obj), value = T)))

ref_set <- markers_v6_super[[coi]] %>% 
  bind_rows(markers_doub_tbl) %>% 
  group_by(subtype) %>% 
  mutate(m = length(gene),
         n = length(rownames(seu_obj))-m,
         join_helper = 1) %>% 
  group_by(subtype, m, n, join_helper) %>%
  nest(ref_set = gene)

hyper_tbl <- test_set %>% 
  left_join(ref_set, by = "join_helper") %>% 
  group_by(cluster, subtype, m, n, k) %>%
  do(q = length(intersect(unlist(.$ref_set), unlist(.$test_set)))) %>%
  mutate(pval = 1-phyper(q = q, m = m, n = n, k = k)) %>%
  ungroup %>%
  mutate(qval = p.adjust(pval, "BH"),
         sig = qval < 0.01)

# hyper_tbl %>% 
#   group_by(subtype) %>% 
#   filter(any(qval < 0.01)) %>%
#   ggplot(aes(subtype, -log10(qval), fill = sig)) +
#   geom_bar(stat = "identity") +
#   facet_wrap(~cluster) +
#   coord_flip()
  
low_rank <- str_detect(unique(hyper_tbl$subtype), "doublet")
subtype_lvl <- c(sort(unique(hyper_tbl$subtype)[!low_rank]), sort(unique(hyper_tbl$subtype)[low_rank]))
  
cluster_label_tbl <- hyper_tbl %>% 
  mutate(subtype = ordered(subtype, levels = subtype_lvl)) %>% 
  arrange(qval, subtype) %>%
  group_by(cluster) %>% 
  slice(1) %>% 
  mutate(subtype = ifelse(sig, as.character(subtype), paste0("unknown_", cluster))) %>% 
  select(cluster, cluster_label = subtype) %>% 
  ungroup %>% 
  mutate(cluster_label = make.unique(cluster_label, sep = "_"))

seu_obj$cluster_label <- unname(deframe(cluster_label_tbl)[as.character(unlist(seu_obj[[paste0("RNA_snn_res.", louvain_resolution)]]))])
plot_data$cluster_label <- seu_obj$cluster_label

marker_sheet <- marker_tbl %>% 
  left_join(cluster_label_tbl, by = "cluster") %>% 
  group_by(cluster_label) %>% 
  filter(!str_detect(gene, "^RPS|^RPL")) %>% 
  slice(1:50) %>% 
  mutate(rank = row_number(-avg_logFC)) %>% 
  select(cluster_label, gene, rank) %>% 
  spread(cluster_label, gene) %>% 
  mutate_all(.funs = helper_f)

formattable::formattable(marker_sheet)

rank	CD4.T.dysfunctional	CD4.T.naive	CD4.T.naive_1	CD4.T.naive_2	CD4.T.reg	CD8.T	CD8.T_1	CD8.T_2	CD8.T_3	Cycling.T.NK	doublet.Fibroblast	doublet.Monocyte	doublet.Plasma.cell	NK.CD56	NK.CD56_1
1	CXCL13	IL7R	KLRB1	KLF2	TNFRSF4	CD8A	IFIT3	HSPA1A	XCL2	STMN1	DCN	HLA-DRA	SOX4	GNLY	FGFBP2
2	NMB	CCR7	IL4I1	CCR7	IL2RA	GZMK	ISG15	HSPA1B	GZMB	MKI67	IGFBP5	CST3	PTCRA	TYROBP	FCGR3A
3	NR3C1	KLF2	IL7R	JUNB	FOXP3	CD8B	MX1	MT1X	CRTAM	TUBA1B	RBP1	CXCL8	MAL	AREG	SPON2
4	FKBP5	EEF1B2	LTB	SELL	CTLA4	CCL4L2	IFIT1	DNAJB1	XCL1	CENPF	CALD1	SPP1	MZB1	KLRC1	PRF1
5	MAF	TPT1	LST1	DUSP1	LTB	CCL4	IFIT2	MT1E	FABP5	TOP2A	C7	S100A9	DNTT	FCER1G	KLRF1
6	IL6ST	EEF1A1	TNFSF13B	FOS	RTKN2	CCL5	RSAD2	HSPA6	TNFRSF9	HIST1H4C	MEG3	S100A8	TFDP2	TRDC	GNLY
7	ITM2A	MAL	CCR6	IL7R	BATF	GZMH	IFI6	HSP90AA1	CXCL13	TUBB	IGFBP4	HLA-DQA1	CD1E	XCL1	KLRD1
8	TSHZ2	TCF7	CTSH	AREG	TNFRSF18	TRGC2	MX2	MT2A	HSP90AB1	HMGB2	ADIRF	CD74	STMN1	KLRD1	NKG7
9	CTLA4	CD40LG	AQP3	EEF1A1	SAT1	ITM2C	IFI44L	HSPH1	PKM	ASPM	RARRES2	LYZ	ARPP21	KRT81	CX3CR1
10	CD40LG	SELL	CCL20	CD69	TBC1D4	CRTAM	ISG20	HSPE1	GAPDH	TYMS	NR2F2	FTL	AC084033.3	XCL2	GZMB
11	PDCD1	GPR183	NFKBIA	BTG2	TIGIT	GZMA	HERC5	HSPD1	NME1	NUSAP1	SELENOP	APOE	CDK6	IGFBP2	PLAC8
12	CD4	LDHB	RORA	GPR183	GADD45A	IFNG	SAMD9L	HSPB1	IFNG	HMGN2	MDK	HLA-DQB1	MAP1A	CLIC3	CLIC3
13	LIMS1	NOSIP	TNFRSF25	PIK3IP1	TNFRSF1B	HLA-DPB1	OAS1	HSP90AB1	RGCC	PCLAF	TAGLN	MARCKS	AC011893.1	IL2RB	PLEK
14	TNFRSF4	SNHG8	CEBPD	CD55	PMAIP1	CXCR6	TNFSF10	DNAJA1	MIR155HG	H2AFZ	MGP	AIF1	GLUL	CEBPD	TYROBP
15	RNF19A	PABPC1	NCR3	DNAJB1	UGP2	KLRG1	STAT1	JUN	TNFRSF18	SMC4	SOX4	C1QB	ADA	TXK	PTGDS
16	CORO1B	NOP53	TNFAIP3	FKBP5	IKZF2	LAG3	EIF2AK2	CACYBP	SRM	PCNA	EGR1	FTH1	AL357060.1	KRT86	EFHD2
17	RBPJ	LEF1	SLC4A10	TSC22D3	TNFRSF9	DTHD1	OAS3	HSPA8	HSPA5	HIST1H1B	IFITM3	BASP1	GRASP	CTSW	FCER1G
18	CPM	LTB	TMIGD2	RACK1	ICOS	CD3G	SAMD9	FKBP4	ENO1	TPX2	C11orf96	C1QA	CD1B	KLRB1	CST7
19	ZBED2	EIF3E	DPP4	LDHB	LINC01943	HLA-DRB1	GBP1	CHORDC1	REL	DUT	CLU	C15orf48	MIR181A1HG	MATK	ADGRG1
20	AC004585.1	RACK1	TPT1	PLAC8	IL32	CST7	XAF1	RGS2	TPI1	UBE2C	SFRP4	FN1	CCDC26	CCL3	GZMH
21	AHI1	NACA	S100A4	CXCR4	SOX4	CD3D	EPSTI1	DNAJB4	NPW	CLSPN	SPARCL1	MNDA	JCHAIN	CD7	CCL3
22	DUSP4	UBA52	MYBL1	SARAF	ARID5B	JAML	IFI44	DNAJA4	EIF5A	SMC2	TCEAL4	G0S2	VIPR2	NKG7	HOPX
23	TOX2	SOCS3	CD40LG	EEF1B2	CD27	PTMS	PLSCR1	ZFAND2A	RBPJ	ATAD2	ADAMTS1	APOC1	ID1	CD63	IGFBP7
24	ICA1	TOMM7	LINC01871	TCF7	BIRC3	HLA-DPA1	USP18	ANXA1	PARK7	CKS1B	TIMP2	SOD2	SOCS2	HOPX	ZEB2
25	ARID5B	JUNB	SPOCK2	TPT1	LAYN	THEMIS	IFI35	PPP1R15A	RGS16	UBE2S	STAR	NPC2	RCAN1	TMIGD2	PRSS23
26	CCDC50	SERINC5	ERN1	FOSB	CORO1B	DUSP2	OASL	FOS	CD160	TMPO	C1R	LST1	CASC15	CMC1	AKR1C3
27	CD84	TMEM123	JAML	SC5D	TYMP	GZMM	MT2A	FOSB	PIM3	PTTG1	LUM	GSN	CLDN5	TNFRSF18	CD247
28	IGFL2	EEF2	FKBP11	NACA	CD4	APOBEC3G	HELZ2	DUSP1	CD82	TUBB4B	CST3	MS4A6A	PFKFB2	KLRC2	MYBL1
29	BATF	FXYD5	IFNGR1	BTG1	DUSP4	LINC02446	TRIM22	SERPINH1	KDM6B	KNL1	SERPINF1	IL1B	GALNT2	SRGAP3	AREG
30	RGS1	TSHZ2	S100A6	CCND3	ENTPD1	TNIP3	CMPK2	TSPYL2	RAN	DEK	WFDC2	PSAP	HES4	GSTP1	S1PR5
31	SRGN	TRABD2A	MGAT4A	NOP53	CTSC	ITGA1	PARP14	UBC	TSPAN17	CENPE	C1S	GRN	MARCKSL1	LAT2	CTSW
32	CH25H	SARAF	EEF1A1	ERAP2	MIR4435-2HG	CD27	LY6E	UBB	NAMPT	HELLS	FHL2	CD83	TP53INP1	GZMB	C1orf21
33	SPOCK2	ANK3	ELK3	ZBTB16	LINC02099	PPP1R14B	NT5C3A	AHSA1	TNFRSF1B	NASP	PEG3	CTSH	APBA2	LINC00996	ABHD17A
34	ZNRF1	AQP3	GPR183	EEF1D	SPOCK2	SLF1	DDX58	NEU1	RANBP1	HMGB1	CEBPD	GLUL	NREP	PRF1	KLF2
35	CHN1	RIPOR2	KIT	PLK3	MAGEH1	CLEC2B	IRF7	DNAJB6	ZBED2	RRM2	NR2F1	MEF2C	TSHR	KLRF1	PTPN12
36	TNFRSF25	AP3M2	LTC4S	PPP1R15A	PHACTR2	IKZF3	RNF213	GADD45B	NDFIP2	BIRC5	TIMP1	CYBB	MME	NCAM1	TTC38
37	CD200	ZFAS1	RORC	ZFP36L2	CARD16	CCL3L1	DDX60L	KLF6	GFOD1	DLGAP5	SERPING1	CTSB	AC002454.1	CXXC5	KLRB1
38	RILPL2	LINC02273	RUNX2	EEF2	S100A4	LINC01871	SAT1	BTG2	GOLIM4	CKS2	AKAP12	CSF3R	CHI3L2	MCTP2	CCL4
39	TNFRSF18	TOB1	ZBTB16	HNRNPA1	STAM	CD84	DDX60	BAG3	MRTO4	FABP5	TSC22D1	CD14	SMIM3	SH2D1B	PTGDR
40	METTL8	EIF4B	FAM241A	VSIR	GLRX	EOMES	PPM1K	TNF	PAICS	H2AFV	TPM2	SGK1	SSBP2	IFITM3	ITGB2
41	SLA	SESN3	IL23R	KLF3	SPATS2L	LYST	PNPT1	ERN1	DCTPP1	H2AFX	BEX3	C1QC	UHRF1	ZNF683	XBP1
42	SMCO4	TNFRSF25	PDE4D	PASK	AC005224.3	TNFSF9	PARP9	JUNB	NOLC1	CDK1	NUPR1	SPI1	LRRC28	ITGA1	CEP78
43	BTLA	NSA2	B3GALT2	NPM1	MAF	COTL1	IFIH1	JUND	EBNA1BP2	GAPDH	DLK1	FCGRT	BCL11A	IFITM2	ARL4C
44	NAP1L4	PASK	EEF1B2	EIF3H	PBXIP1	CD52	SP110	DEDD2	ZNF593	CXCL13	IGFBP7	EGR1	SCAI	CCL5	BIN2
45	MIR155HG	FAU	TLE1	TXNIP	F5	ZNF683	OAS2	CD69	CAMK1	MCM7	COL1A2	FCGR2A	ATP6AP1L	ITGAX	LITAF
46	FYB1	EEF1D	CERK	LDLRAP1	SLAMF1	HCST	C19orf66	CLK1	GEM	RANBP1	CDKN1C	PLAUR	RUFY3	CD38	TRDC
47	PTPN13	LDLRAP1	CFH	CMTM8	BTG3	KIAA1551	STAT2	IER5L	EGR2	CDKN3	FILIP1L	CPVL	CD79A	SAMD3	TXK
48	BIRC3	CTSL	PERP	SCML1	TRAC	GPR174	LAG3	H3F3B	CD72	EZH2	RARRES1	MS4A7	GNA15	SLC16A3	MYOM2
49	SESN3	ITGA6	PLAT	LINC00402	IL1R1	CD3E	LAP3	NR4A1	NOP16	GTSE1	SLC40A1	ALDH2	HHIP-AS1	CAPN12	GZMM
50	AGFG1	PFDN5	KIF5C	RIPK2	DNPH1	PDCD1	APOL6	CXCR4	METTL1	MCM3	TCEAL9	SERPINA1	MYB	CD247	CD300A

write_tsv(marker_sheet, paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_marker_sheet.tsv"))

2 Clusters

2.1 sizes

enframe(sort(table(seu_obj$cluster_label))) %>% 
  mutate(name = ordered(name, levels = rev(name))) %>% 
  ggplot() +
  geom_bar(aes(name, value), stat = "identity") +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5)) +
  labs(y = c("#cells"), x = "cluster")

2.2 UMAP

alpha_lvl <- ifelse(nrow(plot_data) < 20000, 0.2, 0.1)
pt_size <- ifelse(nrow(plot_data) < 20000, 0.2, 0.05)

common_layers_disc <- list(  
  geom_point(size = pt_size, alpha = alpha_lvl),
  NoAxes(),
  guides(color = guide_legend(override.aes = list(size = 2, alpha = 1))),
  labs(color = "")
)

common_layers_cont <- list(  
  geom_point(size = pt_size, alpha = alpha_lvl),
  NoAxes(),
  scale_color_gradientn(colors = viridis(9)),
  guides(color = guide_colorbar())
)

ggplot(plot_data, aes(umapharmony_1, umapharmony_2, color = cluster_label)) + 
  common_layers_disc +
  #facet_wrap(~therapy) +
  ggtitle("Sub cluster")

3 Filtering out doublet and mito high clusters

3.1 UMAP

my_subtypes <- names(clrs$cluster_label[[coi]])

my_subtypes <- c(my_subtypes, unlist(lapply(paste0("_", 1:3), function(x) paste0(my_subtypes, x)))) %>% .[!str_detect(., "doublet")]

cells_to_keep <- colnames(seu_obj)[seu_obj$cluster_label %in% my_subtypes]
seu_obj_sub <- subset(seu_obj, cells = cells_to_keep)
seu_obj_sub <- RunUMAP(seu_obj_sub, dims = 1:50, reduction = "harmony", reduction.name = "umapharmony", reduction.key = "umapharmony_")
seu_obj_sub$cluster_label <- seu_obj$cluster_label[colnames(seu_obj) %in% colnames(seu_obj_sub)]
write_rds(seu_obj_sub, paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_processed_filtered.rds"))
seu_obj_sub <- read_rds(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_processed_filtered.rds"))

plot_data_sub <- as_tibble(FetchData(seu_obj_sub, c(myfeatures, "cluster_label"))) %>% 
  left_join(meta_tbl, by = "sample") %>% 
  mutate(patient_id = str_remove_all(patient_id, "SPECTRUM-OV-"),
         tumor_supersite = ordered(tumor_supersite, levels = names(clrs$tumor_supersite))) %>% 
  mutate(cell_id = colnames(seu_obj_sub),
         cluster_label = ordered(cluster_label, levels = my_subtypes),
         )
  
if (cell_sort == "CD45+") {
plot_data_sub <- filter(plot_data_sub, sort_parameters != "singlet, live, CD45-", !is.na(tumor_supersite))
}

if (cell_sort == "CD45-") {
plot_data_sub <- filter(plot_data_sub, sort_parameters != "singlet, live, CD45+", !is.na(tumor_supersite))
}

ggplot(plot_data_sub, aes(umapharmony_1, umapharmony_2, color = cluster_label)) + 
  common_layers_disc +
  ggtitle("Sub cluster") +
  #facet_wrap(~cluster_label) +
  scale_color_manual(values = clrs$cluster_label[[coi]])

ggplot(plot_data_sub, aes(umapharmony_1, umapharmony_2, color = patient_id_short)) + 
  common_layers_disc +
  ggtitle("Patient") +
  # facet_wrap(~therapy) +
  scale_color_manual(values = clrs$patient_id_short)

ggplot(plot_data_sub, aes(umapharmony_1, umapharmony_2, color = tumor_supersite)) + 
  # geom_point(aes(umapharmony_1, umapharmony_2), 
  #            color = "grey90", size = 0.01, 
  #            data = select(plot_data_sub, -tumor_supersite)) +
  common_layers_disc +
  ggtitle("Site") +
  # facet_wrap(~therapy) +
  scale_color_manual(values = clrs$tumor_supersite)

write_tsv(select(plot_data_sub, cell_id, everything(), -umapharmony_1, -umapharmony_2, -contains("RNA_")), paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_embedding.tsv"))

3.2 add module scores and pathway scores

signature_modules <- read_excel("_data/small/signatures/SPECTRUM v5 sub cluster markers.xlsx", sheet = 2, skip = 1, range = "M2:P100") %>% 
  gather(module, gene) %>% 
  na.omit() %>% 
  group_by(module) %>% 
  do(gene = c(.$gene)) %>% 
  {setNames(.$gene, .$module)}

signature_modules$ISG.module <- c("CCL5", "CXCL10", "IFNA1", "IFNB1", "ISG15", "IFI27L2", "SAMD9L")

## compute expression module scores
for (i in 1:length(signature_modules)) {
  seu_obj_sub <- AddModuleScore(seu_obj_sub, features = signature_modules[i], name = names(signature_modules)[i])
  seu_obj_sub[[names(signature_modules)[i]]] <- seu_obj_sub[[paste0(names(signature_modules)[i], "1")]]
  seu_obj_sub[[paste0(names(signature_modules)[i], "1")]] <- NULL
  print(paste(names(signature_modules)[i], "DONE"))
}

## [1] "CD8.Cytotoxic DONE"
## [1] "CD8.Dysfunctional DONE"
## [1] "CD8.Naive DONE"
## [1] "CD8.Predysfunctional DONE"
## [1] "ISG.module DONE"

## compute progeny scores
progeny_list <- seu_obj_sub@assays$RNA@data[VariableFeatures(seu_obj_sub),] %>% 
  as.matrix %>% 
  progeny %>% 
  as.data.frame %>% 
  as.list

names(progeny_list) <- make.names(paste0(names(progeny_list), ".pathway"))

for (i in 1:length(progeny_list)) {
  seu_obj_sub <- AddMetaData(seu_obj_sub, metadata = progeny_list[[i]], 
                             col.name = names(progeny_list)[i])
}

write_rds(seu_obj_sub, paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_processed_filtered_annotated.rds"))

3.3 marker heatmap

marker_top_tbl <- marker_sheet[,-1] %>% 
  slice(1:10) %>% 
  as.list %>% 
  .[!str_detect(names(.), "doublet")] %>% 
  enframe("cluster_label_x", "gene") %>% 
  unnest(gene)

plot_data_markers <- as_tibble(FetchData(seu_obj_sub, c("cluster_label", myfeatures, unique(marker_top_tbl$gene)))) %>% 
  gather(gene, value, -c(1:(length(myfeatures)+1))) %>% 
  left_join(meta_tbl, by = "sample") %>% 
  mutate(patient_id = str_remove_all(patient_id, "SPECTRUM-OV-"),
         tumor_supersite = ordered(tumor_supersite, levels = names(clrs$tumor_supersite))) %>% 
  mutate(cluster_label = ordered(cluster_label, levels = my_subtypes)) %>% 
  group_by(cluster_label, gene) %>% 
  summarise(value = mean(value, na.rm = T)) %>% 
  group_by(gene) %>% 
  mutate(value = scales::rescale(value)) %>% 
  left_join(marker_top_tbl, by = "gene") %>% 
  mutate(cluster_label_x = ordered(cluster_label_x, levels = rev(names(clrs$cluster_label[[coi]]))))

ggplot(plot_data_markers) +
  geom_tile(aes(gene, cluster_label, fill = value)) +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5)) +
  facet_grid(~cluster_label_x, scales = "free", space = "free") +
  scale_fill_gradientn(colors = viridis(9)) +
  labs(fill = "Scaled\nexpression") +
  theme(aspect.ratio = 1,
        axis.line = element_blank(),
        axis.ticks = element_blank(),
        axis.title = element_blank())

# ggsave(paste0("_fig/002_marker_heatmap_", coi, ".pdf"), width = nrow(marker_top_tbl)/6, height = 5)

3.4 composition

3.4.1 per site

comp_site_tbl <- plot_data_sub %>%
  filter(!is.na(tumor_supersite)) %>% 
  group_by(cluster_label, tumor_supersite) %>%
  tally %>%
  group_by(tumor_supersite) %>%
  mutate(nrel = n/sum(n)*100) %>%
  ungroup

pnrel_site <- ggplot(comp_site_tbl) +
  geom_bar(aes(tumor_supersite, nrel, fill = cluster_label),
           stat = "identity", position = position_stack()) +
  scale_y_continuous(expand = c(0, 0)) +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5)) +
  labs(fill = "Cluster", y = "Fraction [%]", x = "") +
  scale_fill_manual(values = clrs$cluster_label[[coi]])

pnabs_site <- ggplot(comp_site_tbl) +
  geom_bar(aes(tumor_supersite, n, fill = cluster_label),
           stat = "identity", position = position_stack()) +
  scale_y_continuous(expand = c(0, 0)) +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5)) +
  labs(fill = "Cluster", y = "# cells", x = "") +
  scale_fill_manual(values = clrs$cluster_label[[coi]])

plot_grid(pnabs_site, pnrel_site, ncol = 2, align = "h")

# ggsave(paste0("_fig/02_deep_dive_", coi, "_comp_site.pdf"), width = 8, height = 4)

3.4.2 per sample

comp_tbl_sample_sort <- plot_data_sub %>% 
  group_by(tumor_subsite, tumor_supersite, patient_id, consensus_signature, therapy, cluster_label) %>% 
  tally %>% 
  group_by(tumor_subsite, tumor_supersite, patient_id, consensus_signature, therapy) %>% 
  mutate(nrel = n/sum(n)*100,
         nsum = sum(n),
         log10n = log10(n),
         label_supersite = "Site",
         label_mutsig = "Signature",
         label_therapy = "Rx") %>% 
  ungroup %>% 
  arrange(desc(therapy), tumor_supersite) %>% 
  mutate(tumor_subsite_rx = paste0(tumor_subsite, "_", therapy)) %>% 
  mutate(tumor_subsite = ordered(tumor_subsite, levels = unique(tumor_subsite)),
         tumor_subsite_rx = ordered(tumor_subsite_rx, levels = unique(tumor_subsite_rx))) %>% 
  arrange(patient_id) %>% 
  mutate(label_patient_id = ifelse(as.logical(as.numeric(fct_inorder(as.character(patient_id)))%%2), "Patient1", "Patient2"))

sample_id_x_tbl <- plot_data_sub %>% 
  mutate(sort_short_x = cell_sort) %>% 
  distinct(patient_id, sort_short_x, tumor_subsite, therapy, sample) %>% 
  unite(sample_id_x, patient_id, sort_short_x, tumor_subsite, therapy) %>% 
  arrange(sample_id_x)

comp_tbl_sample_sort %>% 
  mutate(sort_short_x = cell_sort) %>% 
  unite(sample_id_x, patient_id, sort_short_x, tumor_subsite_rx) %>% 
  select(sample_id_x, cluster_label, n, nrel, nsum) %>% 
  left_join(sample_id_x_tbl, by = "sample_id_x") %>% 
  write_tsv(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_subtype_compositions.tsv"))

ybreaks <- c(500, 1000, 2000, 4000, 6000, 8000, 10000, 15000, 20000)

max_cells_per_sample <- max(comp_tbl_sample_sort$nsum)
ymaxn <- ybreaks[max_cells_per_sample < ybreaks][1]

comp_plot_wrapper <- function(y = "nrel", switch = NULL) {
  if (y == "nrel") ylab <- paste0("Fraction\nof cells [%]")
  if (y == "n") ylab <- paste0("Number\nof cells")
  p <- ggplot(comp_tbl_sample_sort, 
              aes_string("tumor_subsite_rx", y, fill = "cluster_label")) + 
    facet_grid(~patient_id, space = "free", scales = "free", switch = switch) +
    coord_cartesian(clip = "off") + 
    scale_fill_manual(values = clrs$cluster_label[[coi]]) + 
    theme(axis.text.x = element_blank(),
          axis.title.y = element_text(angle = 0, vjust = 0.5, hjust = 0.5, 
                                      margin = margin(0, -0.4, 0, 0, unit = "npc")),
          axis.ticks.x = element_blank(),
          axis.title.x = element_blank(),
          axis.line.x = element_blank(),
          strip.text.y = element_blank(),
          strip.text.x = element_blank(),
          strip.background.y = element_blank(),
          strip.background.x = element_blank(),
          plot.margin = margin(0, 0, 0, 0, "npc")) + 
    labs(x = "",
         y = ylab) +
    guides(fill = FALSE)
  if (y == "nrel") p <- p + 
    geom_bar(stat = "identity") +
    scale_y_continuous(expand = c(0, 0), 
                       breaks = c(0, 50, 100), 
                       labels = c("0", "50", "100"))
  if (y == "n") p <- p + 
    geom_bar(stat = "identity", position = position_stack()) +
    scale_y_continuous(expand = c(0, 0), 
                       breaks = c(0, ymaxn/2, ymaxn),
                       limits = c(0, ymaxn),
                       labels = c("", ymaxn/2, ymaxn)) +
    expand_limits(y = c(0, ymaxn)) +
    theme(panel.grid.major.y = element_line(linetype = 1, color = "grey90", size = 0.5))
  return(p)
} 

common_label_layers <- list(
  geom_tile(color = "white", size = 0.15),
  theme(axis.text.x = element_blank(),
        axis.ticks = element_blank(),
        axis.title.x = element_blank(),
        axis.line.x = element_blank(),
        strip.text = element_blank(),
        strip.background = element_blank(),
        plot.margin = margin(0, 0, 0, 0, "npc")),
  scale_y_discrete(expand = c(0, 0)),
  labs(y = ""),
  guides(fill = FALSE),
  facet_grid(~patient_id, 
             space = "free", scales = "free")
)

comp_label_site <- ggplot(distinct(comp_tbl_sample_sort, tumor_subsite_rx, therapy, tumor_supersite, label_supersite, patient_id), 
       aes(tumor_subsite_rx, label_supersite, 
           fill = tumor_supersite)) + 
  scale_fill_manual(values = clrs$tumor_supersite) +
  common_label_layers

comp_label_rx <- ggplot(distinct(comp_tbl_sample_sort, tumor_subsite_rx, therapy, tumor_supersite, label_therapy, consensus_signature, patient_id), 
       aes(tumor_subsite_rx, label_therapy, 
           fill = therapy)) + 
  scale_fill_manual(values = c(`post-Rx` = "gold3", `pre-Rx` = "steelblue")) +
  common_label_layers

comp_label_mutsig <- ggplot(distinct(comp_tbl_sample_sort, tumor_subsite_rx, therapy, tumor_supersite, label_mutsig, consensus_signature, patient_id), 
       aes(tumor_subsite_rx, label_mutsig, 
           fill = consensus_signature)) + 
  scale_fill_manual(values = clrs$consensus_signature) +
  common_label_layers

patient_label_tbl <- distinct(comp_tbl_sample_sort, patient_id, .keep_all = T)

comp_label_patient_id <- ggplot(patient_label_tbl, aes(tumor_subsite_rx, label_patient_id)) + 
  scale_fill_manual(values = clrs$consensus_signature) +
  geom_text(aes(tumor_subsite_rx, label_patient_id, label = patient_id)) +
  facet_grid(~patient_id, 
             space = "free", scales = "free") +
  coord_cartesian(clip = "off") + 
  theme_void() +
  theme(strip.text = element_blank(),
        strip.background = element_blank(),
        plot.margin = margin(0, 0, 0, 0, "npc"))

hist_plot_wrapper <- function(x = "nrel") {
  if (x == "nrel") {
    xlab <- paste0("Fraction of cells [%]")
    bw <- 5
  }
  if (x == "log10n") {
    xlab <- paste0("Number of cells")
    bw <- 0.2
  }
  p <- ggplot(comp_tbl_sample_sort) +
    ggridges::geom_density_ridges(
      aes_string(x, "cluster_label", fill = "cluster_label"), color = "black",
      stat = "binline", binwidth = bw, scale = 3) +
    facet_grid(label_supersite~., 
               space = "free", scales = "free") +
    scale_fill_manual(values = clrs$cluster_label[[coi]]) + 
    theme(axis.text.y = element_blank(),
          axis.ticks.y = element_blank(),
          axis.title.y = element_blank(),
          axis.line.y = element_blank(),
          strip.text = element_blank(),
          strip.background = element_blank(),
          plot.margin = margin(0, 0, 0, 0, "npc")) +
    scale_x_continuous(expand = c(0, 0)) +
    scale_y_discrete(expand = c(0, 0)) +
    guides(fill = FALSE) +
    labs(x = xlab)
  if (x == "log10n") p <- p + expand_limits(x = c(0, 3)) + 
    scale_x_continuous(expand = c(0, 0), 
                       labels = function(x) parse(text = paste("10^", x)))
  return(p)
}

pcomp1 <- comp_plot_wrapper("n")
pcomp2 <- comp_plot_wrapper("nrel")

pcomp_grid <- plot_grid(comp_label_patient_id, 
                        pcomp1, pcomp2, 
                        comp_label_site, comp_label_rx, comp_label_mutsig,
                        ncol = 1, align = "v", 
                        rel_heights = c(0.15, 0.33, 0.33, 0.06, 0.06, 0.06))

phist1 <- hist_plot_wrapper("log10n")

pcomp_hist_grid <- ggdraw() +
  draw_plot(pcomp_grid, x = 0.01, y = 0, width = 0.85, height = 1) +
  draw_plot(phist1, x = 0.87, y = 0.05, width = 0.12, height = 0.8)

pcomp_hist_grid

# ggsave(paste0("_fig/02_composition_v6_",coi,".pdf"), pcomp_hist_grid, width = 10, height = 2)

3.4.3 site specific cluster enrichment

comp_tbl_z <- comp_tbl_sample_sort %>% 
  filter(therapy == "pre-Rx",
         !(tumor_supersite %in% c("Ascites", "Other"))) %>% 
  group_by(patient_id, cluster_label) %>% 
  arrange(patient_id, cluster_label, nrel) %>% 
  mutate(rank = row_number(nrel),
         z_rank = scales::rescale(rank)) %>% 
  mutate(mean_nrel = mean(nrel, na.rm = T),
         sd_nrel = sd(nrel, na.rm = T),
         z_nrel = (nrel - mean_nrel) / sd_nrel) %>% 
  ungroup()

ggplot(comp_tbl_z) +
  geom_boxplot(aes(tumor_supersite, z_nrel, color = tumor_supersite), 
               outlier.shape = NA) +
  geom_point(aes(tumor_supersite, z_nrel, color = tumor_supersite), position = position_jitter(), size = 0.1) +
  facet_grid(~cluster_label, scales = "free_x", space = "free_x") +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5),
        aspect.ratio = 5) +
  scale_color_manual(values = clrs$tumor_supersite)

ggplot(comp_tbl_z) +
  geom_boxplot(aes(tumor_supersite, z_rank, color = tumor_supersite), 
               outlier.shape = NA) +
  geom_point(aes(tumor_supersite, z_rank, color = tumor_supersite), position = position_jitter(), size = 0.1) +
  facet_grid(~cluster_label, scales = "free_x", space = "free_x") +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5),
        aspect.ratio = 5) +
  scale_color_manual(values = clrs$tumor_supersite)

3.5 sub cluster CD8 cells

seu_obj_cd8 <- seu_obj_sub %>%
  subset(subset = cluster_label == "CD8.T") %>%
  FindNeighbors(reduction = "harmony", dims = 1:50) %>%
  FindClusters(resolution = 0.2) %>%
  RunUMAP(dims = 1:50, reduction = "harmony", reduction.name = "umapharmony", reduction.key = "umapharmony_")

## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
## 
## Number of nodes: 74288
## Number of edges: 2112884
## 
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.8715
## Number of communities: 18
## Elapsed time: 20 seconds

write_rds(seu_obj_cd8, "/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/CD8.T_processed.rds")
seu_obj_cd8 <- read_rds("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/CD8.T_processed.rds")

marker_tbl <- FindAllMarkers(seu_obj_cd8, only.pos = T)
write_tsv(as_tibble(marker_tbl), "/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/CD8.T_markers.tsv")
marker_tbl <- read_tsv("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/CD8.T_markers.tsv")

## Hypergeometric test --------------------------------------

test_set <- marker_tbl %>% 
  group_by(cluster) %>% 
  filter(!str_detect(gene, "^RPS|^RPL")) %>% 
  slice(1:50) %>% 
  mutate(k = length(cluster)) %>% 
  ungroup %>%
  select(cluster, gene, k) %>% 
  mutate(join_helper = 1) %>% 
  group_by(cluster, join_helper, k) %>%
  nest(test_set = gene)

markers_doub_tbl <- markers_v6 %>% 
  enframe("subtype", "gene") %>% 
  filter(!(subtype %in% unique(c(coi, cell_type_major)))) %>% 
  unnest(gene) %>% 
  group_by(gene) %>% 
  filter(length(gene) == 1) %>% 
  mutate(subtype = paste0("doublet.", subtype)) %>% 
  bind_rows(tibble(subtype = "Mito.high", gene = grep("^MT-", rownames(seu_obj), value = T)))

ref_set <- markers_v6_super[["CD8.T"]] %>% 
  bind_rows(markers_doub_tbl) %>% 
  group_by(subtype) %>% 
  mutate(m = length(gene),
         n = length(rownames(seu_obj))-m,
         join_helper = 1) %>% 
  group_by(subtype, m, n, join_helper) %>%
  nest(ref_set = gene)

hyper_tbl <- test_set %>% 
  left_join(ref_set, by = "join_helper") %>% 
  group_by(cluster, subtype, m, n, k) %>%
  do(q = length(intersect(unlist(.$ref_set), unlist(.$test_set)))) %>%
  mutate(pval = 1-phyper(q = q, m = m, n = n, k = k)) %>%
  ungroup %>%
  mutate(qval = p.adjust(pval, "BH"),
         sig = qval < 0.01)

# hyper_tbl %>% 
#   group_by(subtype) %>% 
#   filter(any(qval < 0.01)) %>%
#   ggplot(aes(subtype, -log10(qval), fill = sig)) +
#   geom_bar(stat = "identity") +
#   facet_wrap(~cluster) +
#   coord_flip()
  
low_rank <- str_detect(unique(hyper_tbl$subtype), "Mito|doublet")
subtype_lvl <- c(sort(unique(hyper_tbl$subtype)[!low_rank]), sort(unique(hyper_tbl$subtype)[low_rank]))
  
cluster_label_tbl <- hyper_tbl %>% 
  mutate(subtype = ordered(subtype, levels = subtype_lvl)) %>% 
  arrange(qval, subtype) %>%
  group_by(cluster) %>% 
  slice(1) %>% 
  mutate(subtype = ifelse(sig, as.character(subtype), paste0("unknown_", cluster))) %>% 
  select(cluster, cluster_label = subtype) %>% 
  ungroup %>% 
  mutate(cluster_label = make.unique(cluster_label, sep = "_"))

marker_sheet <- marker_tbl %>% 
  left_join(cluster_label_tbl, by = "cluster") %>% 
  group_by(cluster_label) %>% 
  filter(!str_detect(gene, "^RPS|^RPL")) %>% 
  slice(1:50) %>% 
  mutate(rank = row_number()) %>% 
  select(cluster_label, gene, rank) %>% 
  spread(cluster_label, gene) %>% 
  mutate_all(.funs = helper_f)

formattable::formattable(marker_sheet)

rank	CD8.T.activated.FOS	CD8.T.activated.FOS_1	CD8.T.dysfunctional.CXCL13	CD8.T.naive	gd.T.cell	Mito.high
1	FOS	CCL4L2	CXCL13	GZMK	GNLY	MALAT1
2	IL7R	IFNG	GZMB	CMC1	TRDC	XIST
3	FOSB	CCL3	CTLA4	ITM2C	XCL1	PTPRC
4	JUN	CCL4	HAVCR2	KLRG1	HOPX	N4BP2L2
5	KLF2	TNF	KRT86	GIMAP7	ZNF683	MT-ND3
6	NFKBIA	CCL3L1	CCL3	CST7	KLRC2	MTRNR2L12
7	JUNB	NFKBID	RBPJ	PLEK	CD7	RNF213
8	DUSP1	EGR2	TNFRSF9	GIMAP4	LGALS1	NEAT1
9	EGR1	TNFSF9	PHLDA1	GZMM	XCL2	DDX17
10	TNFAIP3	NR4A2	MIR155HG	PPP1R14B	LINC02446	MT-ND1
11	TOB1	FOS	TIGIT	SH2D1A	S100A4	MT-CO2
12	CD69	FOSB	ENTPD1	LYAR	CAPG	MT-CO3
13	TSC22D3	NR4A3	TNFRSF18		CD63	AAK1
14	LTB	NR4A1	SRGAP3		IFITM2	IKZF1
15	AC020916.1	EGR3	CCND2		NCR3	MT-CO1
16	ANXA1	FASLG	CD63		S100A6	PNISR
17	CCR7	EGR1	FAM3C		CTSW	NKTR
18	LMNA	KDM6B	PTMS		CD9	FUS
19	TPT1	RILPL2	LAYN		IFITM3	STK4
20	EEF1A1	TAGAP	LAG3		CD52	CD44
21	MT-ND2	CD69	SPRY1		KLRC1	PLCG2
22	CD40LG	GADD45B	GAPDH		TIMP1	ATM
23	AQP3	SLA	NDFIP2		LTB	POLR2J3-1
24	CD55	SRGN	SAMSN1		VIM	ARGLU1
25	PLAC8	ZFP36L1	AKAP5		KLRD1	LUC7L3
26	EEF1B2	XCL2	CLNK		LINC01871	TTC14
27	MT-ND1	DUSP6	GOLIM4		GLUL	MT-ND6
28	PTGER2	KLF6	TNFRSF1B		C1orf162	KANSL1
29	TIMP1	JUND	ID2		TRGC2	RSRP1
30	ZFP36	RASGEF1B	ACP5		SPRY1	IGKC
31	GPR183	MAP3K8	TNFSF4		ACTB	PCSK7
32	BTG2	IER2	ADGRG1		LGALS3	MT-ND4L
33	SOCS3	REL	MYO7A		IL32	LINC00861
34	DPP4	CRTAM	PDCD1		ANKRD28	SF1
35	TXK	PPP1R15A	LINC01871		CTSA	MACF1
36	JUND	AC020916.1	ITGAE		DSTN	MDM4
37	PPP1R15A	SDCBP	DUSP4		GZMB	CCNL1
38	KDM6B	DDX3X	CD7		S100A11	MIAT
39	MT-CYB	CD160	VCAM1		PRMT9	MTRNR2L8
40	KLRB1	XCL1	SNAP47		CKLF	ARID1B
41	SATB1	AC136475.3	IFNG		TRGC1	SORL1
42	MYBL1	FAM53C	RGS1		IKZF2	FTX
43	TCF7	TNFSF14	LINC01943		S100A10	TNRC6B
44	RORA	RGCC	CSF1		ACTG1	PNN
45	AREG	IRF4	SNX9		ID2	CELF2
46	ZFP36L2	DUSP2	HLA-DRA		MATK	HNRNPH1
47	IER2	RNF19A	DGKH		TAGLN2	IFI6
48	TMEM123	NEU1	PRF1		TNFRSF18	PARP14
49	NR4A1	EVI2A	BCL2L11		XBP1	IFI44L
50	IER5	TNFAIP3	FABP5		LDLRAD4	TAPBP

write_tsv(marker_sheet, paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/CD8.T.cell_marker_sheet.tsv"))

seu_obj_cd8$cluster_label <- unname(deframe(cluster_label_tbl)[as.character(unlist(seu_obj_cd8[[paste0("RNA_snn_res.", louvain_resolution)]]))])

seu_obj_sub_sub <- seu_obj_sub

cluster_label_tbl_1 <- as_tibble(cbind(cell_id=colnames(seu_obj_sub), FetchData(seu_obj_sub, c("cluster_label"))))
cluster_label_tbl_2 <- as_tibble(cbind(cell_id=colnames(seu_obj_cd8), FetchData(seu_obj_cd8, c("cluster_label"))))

cluster_label_tbl <- left_join(cluster_label_tbl_1, cluster_label_tbl_2, by = "cell_id") %>%
  mutate(cluster_label = ifelse(is.na(cluster_label.y), cluster_label.x, cluster_label.y))

seu_obj_sub_sub$cluster_label <- cluster_label_tbl$cluster_label

seu_obj_sub_sub <- subset(seu_obj_sub_sub, subset = cluster_label != "doublet.Plasma.cell")
write_rds(seu_obj_sub_sub, "/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/T.cell_processed_filtered_sub.rds")

root_cell <- "SPECTRUM-OV-036_S1_CD45P_PELVIC_PERITONEUM_CGGACACCAACGACTT"
seu_obj_cd8_sub <- subset(seu_obj_cd8, subset = cluster_label != "doublet.Plasma.cell" & cluster_label != "gd.T.cell")
seu_obj_cd8_sub <- subset(seu_obj_cd8_sub, cells = c(root_cell, colnames(seu_obj_cd8_sub)[colnames(seu_obj_cd8_sub)!=root_cell][-1]))
write_rds(seu_obj_sub_sub, "/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/T.cell_processed_filtered_sub.rds")
seu_obj_sub_sub <- read_rds("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/T.cell_processed_filtered_sub.rds")

dc_obj <- DiffusionMap(seu_obj_cd8_sub@reductions$harmony@cell.embeddings, k = 100)
dc_mat <- dc_obj@eigenvectors
colnames(dc_mat) <- paste0("DC_", 1:ncol(dc_mat))
seu_obj_cd8_sub[["DC"]] <- CreateDimReducObject(embeddings = dc_mat, key = "DC_", assay = DefaultAssay(seu_obj_cd8_sub))

write_rds(seu_obj_cd8_sub, "/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/CD8.T_processed_filtered.rds")

dpt_obj <- DPT(dc_obj)
write_rds(dpt_obj, "/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/CD8.T_sub_dpt.rds")
seu_obj_cd8_sub$DPT1 <- dpt_obj$DPT1

write_rds(seu_obj_cd8_sub, "/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/CD8.T_processed_filtered.rds")
seu_obj_cd8_sub <- read_rds("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/CD8.T_processed_filtered.rds")

3.5.1 UMAP

plot_data_sub_sub <- as_tibble(FetchData(seu_obj_sub_sub, c(myfeatures, "cluster_label"))) %>% 
  left_join(meta_tbl, by = "sample") %>% 
  mutate(patient_id = str_remove_all(patient_id, "SPECTRUM-OV-"),
         tumor_supersite = ordered(tumor_supersite, levels = names(clrs$tumor_supersite))) %>% 
  mutate(cell_id = colnames(seu_obj_sub_sub),
         cluster_label = ordered(cluster_label, levels = my_subtypes),
         )
  
if (cell_sort == "CD45+") {
plot_data_sub_sub <- filter(plot_data_sub_sub, sort_parameters != "singlet, live, CD45-", !is.na(tumor_supersite))
}

if (cell_sort == "CD45-") {
plot_data_sub_sub <- filter(plot_data_sub_sub, sort_parameters != "singlet, live, CD45+", !is.na(tumor_supersite))
}

ggplot(plot_data_sub_sub, aes(umapharmony_1, umapharmony_2, color = cluster_label)) + 
  common_layers_disc +
  ggtitle("Sub cluster") +
  #facet_wrap(~cluster_label) +
  scale_color_manual(values = clrs$cluster_label[[coi]])

ggplot(plot_data_sub_sub, aes(umapharmony_1, umapharmony_2, color = patient_id_short)) + 
  common_layers_disc +
  ggtitle("Patient") +
  # facet_wrap(~therapy) +
  scale_color_manual(values = clrs$patient_id_short)

ggplot(plot_data_sub_sub, aes(umapharmony_1, umapharmony_2, color = tumor_supersite)) + 
  # geom_point(aes(umapharmony_1, umapharmony_2), 
  #            color = "grey90", size = 0.01, 
  #            data = select(plot_data_sub_sub, -tumor_supersite)) +
  common_layers_disc +
  ggtitle("Site") +
  # facet_wrap(~therapy) +
  scale_color_manual(values = clrs$tumor_supersite)

write_tsv(select(plot_data_sub_sub, cell_id, everything(), -umapharmony_1, -umapharmony_2, -contains("RNA_")), paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_embedding_sub.tsv"))

3.5.2 UMAP & DC

plot_data_cd8_sub <- as_tibble(FetchData(seu_obj_cd8_sub, c(myfeatures, "cluster_label", "DC_1", "DC_2"))) %>% 
  left_join(meta_tbl, by = "sample") %>% 
  mutate(patient_id = str_remove_all(patient_id, "SPECTRUM-OV-"),
         tumor_supersite = ordered(tumor_supersite, levels = names(clrs$tumor_supersite))) %>% 
  mutate(cell_id = colnames(seu_obj_cd8_sub),
         cluster_label = ordered(cluster_label, levels = my_subtypes),
         )
  
if (cell_sort == "CD45+") {
plot_data_cd8_sub <- filter(plot_data_cd8_sub, sort_parameters != "singlet, live, CD45-", !is.na(tumor_supersite))
}

if (cell_sort == "CD45-") {
plot_data_cd8_sub <- filter(plot_data_cd8_sub, sort_parameters != "singlet, live, CD45+", !is.na(tumor_supersite))
}

ggplot(plot_data_cd8_sub, aes(umapharmony_1, umapharmony_2, color = cluster_label)) + 
  common_layers_disc +
  ggtitle("Sub cluster") +
  #facet_wrap(~cluster_label) +
  scale_color_manual(values = clrs$cluster_label[[coi]])

ggplot(plot_data_cd8_sub, aes(umapharmony_1, umapharmony_2, color = patient_id_short)) + 
  common_layers_disc +
  ggtitle("Patient") +
  # facet_wrap(~therapy) +
  scale_color_manual(values = clrs$patient_id_short)

ggplot(plot_data_cd8_sub, aes(umapharmony_1, umapharmony_2, color = tumor_supersite)) + 
  common_layers_disc +
  ggtitle("Site") +
  # facet_wrap(~therapy) +
  scale_color_manual(values = clrs$tumor_supersite)

ggplot(plot_data_cd8_sub, aes(DC_1, DC_2, color = cluster_label)) + 
  common_layers_disc +
  ggtitle("Sub cluster") +
  #facet_wrap(~cluster_label) +
  scale_color_manual(values = clrs$cluster_label[[coi]])

ggplot(plot_data_cd8_sub, aes(DC_1, DC_2, color = patient_id_short)) + 
  common_layers_disc +
  ggtitle("Patient") +
  # facet_wrap(~therapy) +
  scale_color_manual(values = clrs$patient_id_short)

ggplot(plot_data_cd8_sub, aes(DC_1, DC_2, color = tumor_supersite)) + 
  common_layers_disc +
  ggtitle("Site") +
  # facet_wrap(~therapy) +
  scale_color_manual(values = clrs$tumor_supersite)

4 session info

devtools::session_info()

## ─ Session info ───────────────────────────────────────────────────────────────
##  setting  value                       
##  version  R version 3.6.2 (2019-12-12)
##  os       Debian GNU/Linux 10 (buster)
##  system   x86_64, linux-gnu           
##  ui       X11                         
##  language (EN)                        
##  collate  en_US.UTF-8                 
##  ctype    en_US.UTF-8                 
##  tz       Etc/UTC                     
##  date     2021-01-11                  
## 
## ─ Packages ───────────────────────────────────────────────────────────────────
##  package              * version    date       lib
##  abind                  1.4-5      2016-07-21 [2]
##  ape                    5.3        2019-03-17 [2]
##  assertthat             0.2.1      2019-03-21 [2]
##  backports              1.1.10     2020-09-15 [1]
##  bibtex                 0.4.2.2    2020-01-02 [2]
##  Biobase                2.46.0     2019-10-29 [2]
##  BiocGenerics           0.32.0     2019-10-29 [2]
##  BiocParallel           1.20.1     2019-12-21 [2]
##  bitops                 1.0-6      2013-08-17 [2]
##  boot                   1.3-24     2019-12-20 [3]
##  broom                  0.7.2      2020-10-20 [1]
##  callr                  3.4.2      2020-02-12 [1]
##  car                    3.0-8      2020-05-21 [1]
##  carData                3.0-4      2020-05-22 [1]
##  caTools                1.17.1.4   2020-01-13 [2]
##  cellranger             1.1.0      2016-07-27 [2]
##  class                  7.3-15     2019-01-01 [3]
##  cli                    2.0.2      2020-02-28 [1]
##  cluster                2.1.0      2019-06-19 [3]
##  codetools              0.2-16     2018-12-24 [3]
##  colorblindr          * 0.1.0      2020-01-13 [2]
##  colorspace           * 1.4-2      2019-12-29 [2]
##  cowplot              * 1.0.0      2019-07-11 [2]
##  crayon                 1.3.4      2017-09-16 [1]
##  curl                   4.3        2019-12-02 [2]
##  data.table             1.12.8     2019-12-09 [2]
##  DBI                    1.1.0      2019-12-15 [2]
##  dbplyr                 2.0.0      2020-11-03 [1]
##  DelayedArray           0.12.2     2020-01-06 [2]
##  DEoptimR               1.0-8      2016-11-19 [1]
##  desc                   1.2.0      2018-05-01 [2]
##  destiny              * 3.0.1      2020-01-16 [1]
##  devtools               2.2.1      2019-09-24 [2]
##  digest                 0.6.25     2020-02-23 [1]
##  dplyr                * 1.0.2      2020-08-18 [1]
##  e1071                  1.7-3      2019-11-26 [1]
##  ellipsis               0.3.1      2020-05-15 [1]
##  evaluate               0.14       2019-05-28 [2]
##  fansi                  0.4.1      2020-01-08 [2]
##  farver                 2.0.3      2020-01-16 [1]
##  fitdistrplus           1.0-14     2019-01-23 [2]
##  forcats              * 0.5.0      2020-03-01 [1]
##  foreign                0.8-74     2019-12-26 [3]
##  formattable            0.2.0.1    2016-08-05 [1]
##  fs                     1.5.0      2020-07-31 [1]
##  future                 1.15.1     2019-11-25 [2]
##  future.apply           1.4.0      2020-01-07 [2]
##  gbRd                   0.4-11     2012-10-01 [2]
##  gdata                  2.18.0     2017-06-06 [2]
##  generics               0.0.2      2018-11-29 [2]
##  GenomeInfoDb           1.22.0     2019-10-29 [2]
##  GenomeInfoDbData       1.2.2      2020-01-14 [2]
##  GenomicRanges          1.38.0     2019-10-29 [2]
##  ggplot.multistats      1.0.0      2019-10-28 [1]
##  ggplot2              * 3.3.2      2020-06-19 [1]
##  ggrepel                0.8.1      2019-05-07 [2]
##  ggridges               0.5.2      2020-01-12 [2]
##  ggthemes               4.2.0      2019-05-13 [1]
##  globals                0.12.5     2019-12-07 [2]
##  glue                   1.3.2      2020-03-12 [1]
##  gplots                 3.0.1.2    2020-01-11 [2]
##  gridExtra              2.3        2017-09-09 [2]
##  gtable                 0.3.0      2019-03-25 [2]
##  gtools                 3.8.1      2018-06-26 [2]
##  haven                  2.3.1      2020-06-01 [1]
##  hexbin                 1.28.0     2019-11-11 [2]
##  hms                    0.5.3      2020-01-08 [1]
##  htmltools              0.4.0      2019-10-04 [2]
##  htmlwidgets            1.5.1      2019-10-08 [2]
##  httr                   1.4.2      2020-07-20 [1]
##  ica                    1.0-2      2018-05-24 [2]
##  igraph                 1.2.5      2020-03-19 [1]
##  IRanges                2.20.2     2020-01-13 [2]
##  irlba                  2.3.3      2019-02-05 [2]
##  jsonlite               1.7.1      2020-09-07 [1]
##  KernSmooth             2.23-16    2019-10-15 [3]
##  knitr                  1.26       2019-11-12 [2]
##  knn.covertree          1.0        2019-10-28 [1]
##  labeling               0.3        2014-08-23 [2]
##  laeken                 0.5.1      2020-02-05 [1]
##  lattice                0.20-38    2018-11-04 [3]
##  lazyeval               0.2.2      2019-03-15 [2]
##  leiden                 0.3.1      2019-07-23 [2]
##  lifecycle              0.2.0      2020-03-06 [1]
##  listenv                0.8.0      2019-12-05 [2]
##  lmtest                 0.9-37     2019-04-30 [2]
##  lsei                   1.2-0      2017-10-23 [2]
##  lubridate              1.7.9.2    2020-11-13 [1]
##  magrittr             * 2.0.1      2020-11-17 [1]
##  MASS                   7.3-51.5   2019-12-20 [3]
##  Matrix                 1.2-18     2019-11-27 [3]
##  matrixStats            0.56.0     2020-03-13 [1]
##  memoise                1.1.0      2017-04-21 [2]
##  metap                  1.2        2019-12-08 [2]
##  mnormt                 1.5-5      2016-10-15 [2]
##  modelr                 0.1.8      2020-05-19 [1]
##  multcomp               1.4-12     2020-01-10 [2]
##  multtest               2.42.0     2019-10-29 [2]
##  munsell                0.5.0      2018-06-12 [2]
##  mutoss                 0.1-12     2017-12-04 [2]
##  mvtnorm                1.0-12     2020-01-09 [2]
##  nlme                   3.1-143    2019-12-10 [3]
##  nnet                   7.3-12     2016-02-02 [3]
##  npsurv                 0.4-0      2017-10-14 [2]
##  numDeriv               2016.8-1.1 2019-06-06 [2]
##  openxlsx               4.1.5      2020-05-06 [1]
##  pbapply                1.4-2      2019-08-31 [2]
##  pcaMethods             1.78.0     2019-10-29 [2]
##  pillar                 1.4.6      2020-07-10 [1]
##  pkgbuild               1.0.6      2019-10-09 [2]
##  pkgconfig              2.0.3      2019-09-22 [1]
##  pkgload                1.0.2      2018-10-29 [2]
##  plotly                 4.9.1      2019-11-07 [2]
##  plotrix                3.7-7      2019-12-05 [2]
##  plyr                   1.8.5      2019-12-10 [2]
##  png                    0.1-7      2013-12-03 [2]
##  prettyunits            1.1.1      2020-01-24 [1]
##  processx               3.4.2      2020-02-09 [1]
##  progeny              * 1.11.3     2020-10-22 [1]
##  proxy                  0.4-24     2020-04-25 [1]
##  ps                     1.3.2      2020-02-13 [1]
##  purrr                * 0.3.4      2020-04-17 [1]
##  R.methodsS3            1.7.1      2016-02-16 [2]
##  R.oo                   1.23.0     2019-11-03 [2]
##  R.utils                2.9.2      2019-12-08 [2]
##  R6                     2.4.1      2019-11-12 [1]
##  ranger                 0.12.1     2020-01-10 [1]
##  RANN                   2.6.1      2019-01-08 [2]
##  rappdirs               0.3.1      2016-03-28 [2]
##  RColorBrewer           1.1-2      2014-12-07 [2]
##  Rcpp                   1.0.4      2020-03-17 [1]
##  RcppAnnoy              0.0.16     2020-03-08 [1]
##  RcppEigen              0.3.3.7.0  2019-11-16 [2]
##  RcppHNSW               0.2.0      2019-09-20 [2]
##  RcppParallel           4.4.4      2019-09-27 [2]
##  RCurl                  1.98-1.1   2020-01-19 [1]
##  Rdpack                 0.11-1     2019-12-14 [2]
##  readr                * 1.4.0      2020-10-05 [1]
##  readxl               * 1.3.1      2019-03-13 [2]
##  rematch                1.0.1      2016-04-21 [2]
##  remotes                2.1.0      2019-06-24 [2]
##  reprex                 0.3.0      2019-05-16 [2]
##  reshape2               1.4.3      2017-12-11 [2]
##  reticulate             1.14       2019-12-17 [2]
##  rio                    0.5.16     2018-11-26 [1]
##  rlang                  0.4.8      2020-10-08 [1]
##  rmarkdown              2.0        2019-12-12 [2]
##  robustbase             0.93-6     2020-03-23 [1]
##  ROCR                   1.0-7      2015-03-26 [2]
##  rprojroot              1.3-2      2018-01-03 [2]
##  RSpectra               0.16-0     2019-12-01 [2]
##  rstudioapi             0.11       2020-02-07 [1]
##  rsvd                   1.0.3      2020-02-17 [1]
##  Rtsne                  0.15       2018-11-10 [2]
##  rvest                  0.3.6      2020-07-25 [1]
##  S4Vectors              0.24.2     2020-01-13 [2]
##  sandwich               2.5-1      2019-04-06 [2]
##  scales                 1.1.0      2019-11-18 [2]
##  scatterplot3d          0.3-41     2018-03-14 [1]
##  sctransform            0.2.1      2019-12-17 [2]
##  SDMTools               1.1-221.2  2019-11-30 [2]
##  sessioninfo            1.1.1      2018-11-05 [2]
##  Seurat               * 3.1.2      2019-12-12 [2]
##  SingleCellExperiment   1.8.0      2019-10-29 [2]
##  smoother               1.1        2015-04-16 [1]
##  sn                     1.5-4      2019-05-14 [2]
##  sp                     1.4-2      2020-05-20 [1]
##  stringi                1.5.3      2020-09-09 [1]
##  stringr              * 1.4.0      2019-02-10 [1]
##  SummarizedExperiment   1.16.1     2019-12-19 [2]
##  survival               3.1-8      2019-12-03 [3]
##  testthat               2.3.2      2020-03-02 [1]
##  TFisher                0.2.0      2018-03-21 [2]
##  TH.data                1.0-10     2019-01-21 [2]
##  tibble               * 3.0.4      2020-10-12 [1]
##  tidyr                * 1.1.2      2020-08-27 [1]
##  tidyselect             1.1.0      2020-05-11 [1]
##  tidyverse            * 1.3.0      2019-11-21 [2]
##  tsne                   0.1-3      2016-07-15 [2]
##  TTR                    0.23-6     2019-12-15 [1]
##  usethis                1.5.1      2019-07-04 [2]
##  uwot                   0.1.5      2019-12-04 [2]
##  vcd                    1.4-7      2020-04-02 [1]
##  vctrs                  0.3.5      2020-11-17 [1]
##  VIM                    6.0.0      2020-05-08 [1]
##  viridis              * 0.5.1      2018-03-29 [2]
##  viridisLite          * 0.3.0      2018-02-01 [2]
##  withr                  2.3.0      2020-09-22 [1]
##  xfun                   0.12       2020-01-13 [2]
##  xml2                   1.3.2      2020-04-23 [1]
##  xts                    0.12-0     2020-01-19 [1]
##  XVector                0.26.0     2019-10-29 [2]
##  yaml                   2.2.1      2020-02-01 [1]
##  zip                    2.0.4      2019-09-01 [1]
##  zlibbioc               1.32.0     2019-10-29 [2]
##  zoo                    1.8-7      2020-01-10 [2]
##  source                                 
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  Bioconductor                           
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Github (clauswilke/colorblindr@1ac3d4d)
##  R-Forge (R 3.6.2)                      
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  Bioconductor                           
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Github (saezlab/progeny@94bea1c)       
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.3)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
## 
## [1] /home/uhlitzf/R/lib
## [2] /usr/local/lib/R/site-library
## [3] /usr/local/lib/R/library

MSK SPECTRUM freeze major subset deep dives

Florian Uhlitz

2021-01-10

1 Cluster markers

1.1 Major T.super markers for cell assign

1.2 Subtype currated markers

1.3 Subtype cluster markers

2 Clusters

2.1 sizes

2.2 UMAP

3 Filtering out doublet and mito high clusters

3.1 UMAP

3.2 add module scores and pathway scores

3.3 marker heatmap

3.4 composition

3.4.1 per site

3.4.2 per sample

3.4.3 site specific cluster enrichment

3.5 sub cluster CD8 cells

3.5.1 UMAP

3.5.2 UMAP & DC

4 session info